Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 1 of 85 PageID #: 69497 EXHIBIT 4

Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 1 of 85 PageID #: 69497 EXHIBIT 4

Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 2 of 85 PageID #: 69498 UNITED STATES DISTRICT COURT EASTERN DISTRICT OF NEW YORK IN RE PAYMENT CARD INTERCHANGE FEE AND MERCHANT DISCOUNT ANTITRUST LITIGATION No. 05-MD-1720 (IG) (JO) This Document Applies to: All Cases. REPLY DECLARATION OF ALAN S. FRANKEL, Ph.D. RELATING TO THE PROPOSED CLASS SETTLEMENT

Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 3 of 85 PageID #: 69499 1. Introduction 1 2. Professor Hausman's Claim That The Proposed Settlement Will Not "Eliminate" The Defendants' Market Power Or Establish "Competitive" Interchange Fees Is Irrelevant To My Conclusions 4 3. Professor Hausman's Analysis Of No-Surcharge Rules Is Inconsistent With His Analysis Of Honor-All-Cards Rules 5 4. The Evidence Shows That The Ability To Surcharge Is Valuable To Merchants 11 4.1 Many Merchants Surcharge When Merchant Fees Are At Levels Prevailing For MasterCard And Visa Credit Card Acceptance In The United States 13 4.1.1 In Australia, Merchants Commonly Surcharge Cards That Carry Merchant Fees Similar To The Networks' Cards In The United States 14 4.1.2 Professor Hausman Misstates The Facts About The Level Of Interchange Fees and Merchant Fees In The United Kingdom 20 4.2 Surcharging Causes Many Customers To Switch To Non-Surcharged Payment Options 25 4.3 The Ability To Surcharge Constrains The Level Of Merchant Fees When They Are At The Level Prevailing In The United State 26 4.3.1 Australian Evidence Supports My Conclusion That The Ability To Surcharge Constrains High Merchant Card Acceptance Fees 28 4.3.1.1 Surcharging Contributed To Merchant Fee Reductions By American Express And Diners Club 29 4.3.1.2 Whether Surcharges Are Sufficient To Result In Interchange Fees Of 0.50% or 0.30% Is Irrelevant To Whether They Will Benefit U.S. Merchants Now Paying Many Times That Amount 38 4.3.2 Countries Where Surcharging Is Permitted Have Below-Average Interchange Fee Rates 40 4.4 State Statutes And American Express's Policy Reduce But Do Not Eliminate The Benefits To Merchants From The Ability For Merchants To Surcharge MasterCard and Visa Credit Card Transactions 41 4.4.1 State Statutes 42 4.4.2 American Express's Policy 44 5. Professor Hausman and Professor Weisbach Misunderstand The Purpose Of My Illustrative Projections Of Benefits To Merchants From The Ability To Surcharge 45 6. Conclusion 47 Appendix A: Detailed Responses To Professors Hausman And Weisbach Regarding Projected Merchant Benefits 48 A.1 Professor Hausman's Criticisms 49 A.2 Professor Weisbach's Criticisms 60

Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 4 of 85 PageID #: 69500 1. Introduction 1. I submitted a declaration' ("Frankel Declaration") concerning elements of the proposed class settlement 2 ("Proposed Settlement") on April 10, 2013. I previously submitted a report ("Frankel Report") and a rebuttal report ("Frankel Rebuttal Report") concerning liability and damages issues in this case. 3 2. In the Frankel Declaration, I explained the bases for my three main opinions: (1) merchants benefit from the ability to steer customers to less costly payment methods, an ability that is enhanced by the Proposed Settlement; (2) the relief provided by the Dodd-Frank Act and the DOJ Settlement is helpful to merchants, and the Proposed Settlement's preservation of that relief is therefore valuable; and (3) the ability to surcharge credit card transactions, refuse to accept cards (or surcharge them) at particular banners, and the requirement for MasterCard and Visa (together, "the Networks") to negotiate with buying 1 2 3 Declaration of Alan S. Frankel, Ph.D. Relating to the Proposed Class Settlement, April 10, 2013 ("Frankel Declaration"). In Re Payment Card Interchange Fee and Merchant Discount Antitrust Litigation, No. 05-MDL-1720 (1G)(10) Definitive Class Settlement Agreement. Report of Alan S. Frankel, Ph.D., July 9, 2009 ("Frankel Report"); Rebuttal Report of Alan S. Frankel, Ph.D., June 22, 2010 ("Frankel Rebuttal Report"). My background and qualifications regarding the issues in this case are also described in those reports. Since submitting my Rebuttal Report, I have submitted two reports and testified as an expert on behalf of Canada's Commissioner of Competition in a case against MasterCard and Visa regarding certain of those networks' merchant rules. I also spoke about related issues at conferences hosted by the Federal Reserve Bank of Kansas City in March 2012, and The Clearing House Association in New York in November 2012. I attach my current CV as Exhibit 1. In preparing this declaration I drew on my extensive research into these issues as described in my earlier reports in this case, the materials cited therein, and in my published writings. I attach a list of materials that I specifically relied on in preparing this declaration in Exhibit 2. 1

Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 5 of 85 PageID #: 69501 groups of merchants in good faith will generate significant benefits for merchants by providing additional, effective means of constraining the Networks' market power. 4 3. Professor Jerry Hausman has filed a report 5 ("Hausman Report") addressing certain liability issues, the Proposed Settlement, and my conclusions regarding the likely economic effects of certain aspects of the Proposed Settlement, particularly the reform of MasterCard and Visa no-surcharge rules. Professor Hausman also describes what he believes would be more effective relief for merchants (and presumably believes would be a better settlement) elimination of the Defendants' "honor-all-cards" rules. 6 4. In Part 2, I address the standard by which I evaluated the Proposed Settlement. Professor Hausman evaluates the Proposed Settlement with reference to a standard of whether it will "eliminate [the Networks] market power" and prevent interchange fees from being set "above the competitive level."' He does not explain what he means by "the competitive level" of interchange fees. 8 In any event, I did not evaluate the effects of the Proposed Settlement with respect to a benchmark of whether it will eliminate the Defendants' market power and will 4 5 6 7 8 Frankel Declaration, (11 1114-6. Report of Professor Jerry Hausman, In re Payment Card Interchange Fee and Merchant Discount Antitrust Litigation, May 28, 2013 ("Hausman Report"). Professor Hausman appears to be advocating elimination of the Defendants' "honor-all-cards" rules as an alternative to the settlement's modification of the Defendants' no-surcharge rules, but his report is somewhat ambiguous on this point. If he were advocating elimination of the Defendants' "honor-all-cards" rules in addition to the relief contained in the Proposed Settlement, that would not alter my analysis or conclusions in a substantive way. Hausman Report, 1053. MasterCard's honor-all-cards rules require that a merchant that accepts any MasterCard branded credit card accept all such credit cards, and merchants that accept any MasterCard branded debit card accept all such debit cards. Visa also has a similar honor-all-cards rule. To the extent that he defines the "competitive" interchange fees to be those which would result from the absence of the honor-all-cards rule, it would be merely a tautology for him to assert that his proposal (to eliminate the honor-all-cards rule) would result in the competitive level of interchange fees. 2

Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 6 of 85 PageID #: 69502 result in perfectly competitive pricing to merchants for acceptance of MasterCard and Visa credit cards but rather "whether that relief will benefit the members of the class by reducing the cost to merchants to accept MasterCard or Visa credit card payments from their customers." 9 5. In Part 3, I review Professor Hausman's analysis of the "business stealing" effect, how that effect contributed to market power, and how it can make surcharging more difficult and less likely for some merchants. I explain, however, that surcharging generates less of a business-stealing effect than the all-or-nothing choice that merchants have had until now, so surcharging can be expected to generate enhanced competition. I also explain why surcharging is likely to generate less of a business-stealing problem for merchants than the remedy that Professor Hausman would prefer, the elimination of the Networks' "honor-all-cards" rules. 6. In Part 4, I show that the available economic evidence supports my conclusion that the ability to surcharge is valuable to merchants. In particular, with interchange fees as high as they are in the United States, merchants are likely to surcharge; consumers react to surcharging by switching in significant numbers to alternative non-surcharged cards and payment methods; and the ability to surcharge constrains the level of fees that card networks charge to merchants. Professor Hausman disagrees that merchants will surcharge or that surcharges will constrain the level of interchange fees, but I explain that his conclusions rest on an incomplete and flawed review of the evidence. I also discuss Professor Hausman's opinions concerning state surcharge-related statutes and American Express's policy. Professor Hausman believes that those factors eliminate any benefits that otherwise might be derived from the 9 Frankel Declaration, 111. 3

Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 7 of 85 PageID #: 69503 elimination of the Networks' no-surcharge rules. I disagree, and I explain why those factors affect but do not eliminate the benefits to merchants from the Proposed Settlement. 7. In Part 5, I respond to criticisms by Professor Hausman and by Professor Michael Weisbach of my projections which show that even modest amounts of surcharging and responses to surcharging can generate billions of dollars in savings to merchants. I explain that their criticisms do not alter my conclusion or the usefulness of my projections for illustrating the basis for my conclusion. I respond in more detail to their criticisms in Appendix A. 8. I end in Part 6 with a brief conclusion. 2. Professor Hausman's Claim That The Proposed Settlement Will Not "Eliminate" The Defendants' Market Power Or Establish "Competitive" Interchange Fees Is Irrelevant To My Conclusions 9. The standard by which the reasonableness of a settlement is evaluated is a legal issue, not an economic one. Professor Hausman recommends reforms to MasterCard and Visa honor-all-cards rules that he conjectures (in combination with technological advances) could result in the "elimination of [those Networks'] market power in the future." 1 In contrast to his opinion concerning the effects of the honor-all-cards rule, Professor Hausman predicts that "modification of the no surcharge rule would not eliminate [the Networks'] market power, and interchange would continue to be above the competitive level." 11 10. I was neither asked, nor did I address, whether the Proposed Settlement would eliminate MasterCard's or Visa's market power altogether. Instead, I explained in my previous 1 Hausman Report, 1110. 31 Hausman Report, 1153. 4

Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 8 of 85 PageID #: 69504 declaration that my assignment was "to evaluate the likely economic impact of the relief with respect to merchant rules contained in the [Proposed Settlement], and, in particular, whether that relief will benefit the members of the class by reducing the cost to merchants to accept MasterCard or Visa credit card payments from their customers." 12 I concluded that the Proposed Settlement will generate benefits for merchants and constrain MasterCard's and Visa's market power. When evaluating the value of the Proposed Settlement to merchants, I concluded that it was reasonable to assume that the additional constraints could lead to a reduction of roughly 1-4 basis points (0.01% - 0.04%) per year in the average level of interchange fees, not that competitive forces would speedily drive credit card merchant discount rates to competitive levels. Indeed, MasterCard and Visa interchange fees have increased significantly over the past fifteen years. Slowing and reversing that trend would represent a substantial benefit to merchants relative to what they otherwise would pay even if merchants do not quickly obtain fully competitive prices for card acceptance services. Professor Hausman's straw man argument and extreme standard miss the point of my analysis. 3. Professor Hausman's Analysis Of No-Surcharge Rules Is Inconsistent With His Analysis Of Honor-All-Cards Rules 11. In this part, I review what Professor Hausman refers to as the "business stealing" effect, how it contributes to the Networks' market power, and how surcharging generates less of a business-stealing effect than declining a Network's cards altogether and thereby reduces a Network's market power. Professor Hausman contends that eliminating the Networks' honorall-cards rules, to permit merchants to selectively accept credit cards by some issuers but not 12 Frankel Declaration, Ifil.. 5

Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 9 of 85 PageID #: 69505 others, would reduce the Networks' market power. He fails to note that surcharging is likely to involve less of a business-stealing effect than selective acceptance and is therefore more likely, not less, to constrain the Networks' market power. 12. Professor Hausman agrees with me about many of the economic features of the marketplace. Although his terminology differs somewhat from mine, Professor Hausman agrees that MasterCard and Visa each possesses significant market power with respect to general purpose card network services. 13 One of the sources for that market power is what Professor Hausman refers to as a "business stealing" effect that causes merchants adopting a cost-reducing strategy to lose sales and profits to merchants that do not adopt such a strategy. To explain, with the Anti-Steering Rules in place, a merchant's main option with respect to MasterCard or Visa credit cards to avoid paying high fees was to discontinue to accept one (or both) brands. For example, suppose that Visa increased its merchant fees so that they cost significantly more to a merchant than MasterCard credit cards. If the merchant considered whether to stop accepting Visa credit cards, it would have to take into consideration two main issues. First, some of the merchant's customers who prefer to use a Visa credit card would likely continue to patronize the merchant but use another credit card, debit card, checks, or cash when completing transactions. But other customers might switch to a competing 13 Frankel Report, Part 4; Hausman Report, MO, 27. Unlike me, Professor Hausman contends that all debit cards are in a single relevant market, and that Visa, but apparently not MasterCard, possesses market power in that market. Hausman Report, 1j41. I have explained in detail in my earlier reports the bases for my opinions regarding market definition and market power. Although Professor Hausman's reasoning concerning the relevant debit card market(s) is flawed, and MasterCard, like Visa, possesses market power over the acceptance of signature debit transaction, issues concerning market definition and market power are not directly relevant to the subjects I addressed in the Frankel Declaration. I note, however, that the Proposed Settlement provides relief with respect to MasterCard debit cards. In particular, should current Federal Reserve regulation of MasterCard debit card interchange fees terminate merchants will have the option to surcharge MasterCard debit card transactions under the terms of the Proposed Settlement. 6

69506 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 10 of 85 PageID #: merchant that continues to accept Visa credit cards. Merchants that continue to accept Visa cards "steal business" from those that do not. 14 Professor Hausman and I agree that, for many merchants, losing sales can be relatively costly because merchants tend to earn significant incremental profits on those sales. 15 Declining to accept Visa credit cards would only be profitable if the savings from lower payment costs exceed the lost profits on sales to customers who defect to competing merchants. For most typical merchants, this analysis leads to a conclusion that refusing to accept Visa (or MasterCard) credit cards is not an economically viable strategy. 13. Professor Hausman argues that the same business-stealing effect that deters merchants from dropping acceptance of either MasterCard or Visa credit cards will also deter merchants from surcharging credit card transactions. 16 Using a similar analysis as described above for refusing to accept a brand of cards altogether, Professor Hausman notes that if enough customers defect from a merchant that surcharges the brand carried on the customers' preferred credit card to merchants that do not surcharge, then surcharging will not be profitable. 14. Professor Hausman's analysis is incomplete. I explained in my Declaration that the business-stealing effect can create a "first mover disadvantage" for some merchants that 14 15 16 Merchants would be better off if they could coordinate their acceptance decisions. Without coordination, all end up accepting the card and none gain this "business-stealing" benefit relative to other merchants. While bank interchange fee income was determined on behalf of all banks collectively through MasterCard and Visa, merchants could not coordinate. The Proposed Settlement includes a provision designed to facilitate merchant coordination in negotiating with MasterCard and Visa. Frankel Declaration, Part 4.5. Hausman Report, 1P1144, 66. Hausman Report, 111144, 76. 7

69507 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 11 of 85 PageID #: may make them reluctant to be among the first to surcharge credit cards. 17 There is an important distinction between the effects on a merchant of dropping a brand of credit cards altogether and accepting that brand of credit cards with a surcharge. As I explained in my initial report, lailthough merchants are generally reluctant to stop accepting cards because of the possibility of lost sales, there is less risk of such lost sales from surcharging." 18 If a merchant drops acceptance of a Network's cards, the merchant cannot accept any transactions under any terms from customers using that Network. If the customer does not carry a credit card that the merchant continues to accept, or the customer has strong preferences to use a particular card, the customer may be likely to patronize a different, less preferred merchant (in order to continue using the cardholder's preferred, or only, credit card). If, on the other hand, the merchant continues to accept the Network's cards with a surcharge, then some customers who prefer to patronize the surcharging merchant will continue to do so and pay the surcharge (in which case the merchant also obtains the additional surcharge revenue to defray the cost of card acceptance) rather than switch to another merchant. By surcharging costly credit card transactions, the merchant can also profitably offer lower posted prices to customers who use lower cost payment brands and methods. 19 Lower posted prices will tend to increase the merchant's sales and profits on sales to users of lower cost payment methods such as debit cards and cash. 17 18 19 Frankel Declaration, Part 4.1.1. Frankel Report, 11184. Frankel Rebuttal Report, Part 4.6.2. 8

69508 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 12 of 85 PageID #: 15. Surcharging the use of a card does not eliminate the business-stealing effect, but it reduces that effect relative to not accepting a card thereby increasing the Networks' loss of transactions as they increase their merchant fees i.e., surcharging reduces the Networks' market power. The ability to surcharge thus creates a deterrent to the Networks from increasing their fees that does not exist when merchants' only choice is whether or not to accept each credit card brand. 16. Professor Hausman agrees that MasterCard and Visa have maintained their market power through use of various "restrictive policies." 20 Professor Hausman further agrees that merchants would benefit further by obtaining additional means to intensify competition faced by MasterCard and Visa. But Professor Hausman focuses on the effects of the hypothetical elimination of the Networks' honor-all-cards rules. Professor Hausman argues that elimination of the Networks' honor-all-cards rules, so that a merchant could selectively accept a Network's cards issued by some banks but not others, or the Network's standard credit cards but not the Network's "premium" credit cards, would lead to lower interchange fees. 21 17. Although some merchants might selectively accept only some of the Networks' cards (if they were permitted to do so), the ability to surcharge is more likely to be used or credibly threatened and therefore a more effective steering tool for merchants. This is because, although Professor Hausman addresses the business-stealing problem that enhances the Network's market power with respect to refusal of acceptance or surcharging, he neglects 20 21 Hausman Report, 1147. Hausman Report, $1148-52. 9

69509 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 13 of 85 PageID #: to consider business stealing when he discusses selective acceptance. Professor Hausman reasons that if there were no honor-all-cards rules, an individual merchant could refuse to accept the cards from an individual Issuer unless that Issuer agreed to accept lower interchange fee rates. In Professor Hausman's opinion, this would be sufficient to reduce interchange fees although he does not say whether it would attain his own standard of establishing "competitive" interchange fees. But the "business-stealing" problem confronting merchants from surcharging is also present were a merchant to consider dropping acceptance of a particular Issuer's credit cards following the hypothetical elimination of the honor-all-cards rule. Indeed, the problem could be worse because the likelihood of a given customer leaving without making a purchase would be higher if the merchant did not accept a card carried by the customer than if the merchant accepted the card with a surcharge. If the customer lacks any other payment method, then the customer is certain to leave if the merchant does not accept the card, while with a surcharge some of those customers will stay and pay the surcharge. Similarly, some customers will choose to leave even if they have alternative payment mechanisms if they cannot use their preferred card, but, again, some of those customers would remain and pay the surcharge if that were an option. 18. Ultimately, using any steering strategy, a merchant must compare the lost profits from customers who have no accepted means of payment available (or who defect to other merchants to use a preferred credit card) to the savings resulting from customers who continue to patronize the merchants at a lower cost and additional customers attracted to the merchant due to its ability to offer lower prices. Some merchants might conclude that selective acceptance is a viable strategy. But because surcharging offers an additional option to 10

69510 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 14 of 85 PageID #: customers to continue using their preferred card with a surcharge it can be expected to result in a reduced probability of losing a sale to an affected customer along with the resulting profit from that sale. If, as Professor Hausman contends, the ability to adopt selective acceptance would be likely to benefit merchants, then so will the ability to surcharge. 22 4. The Evidence Shows That The Ability To Surcharge Is Valuable To Merchants 19. In my initial declaration, I reached the following conclusions relating to how merchant surcharging operates as a competitive tool to reduce costs and constrain the Networks' market power: a. Merchants are more likely to apply surcharges to a brand or type of credit card transactions as the level of merchant fees for accepting those cards increases. 23 b. Consumers do not like to pay surcharges and often switch to other credit card brands or payment methods to avoid paying surcharges. 24 Merchants that surcharge thus benefit directly by shifting a significant number of transactions to lower cost payments and by recovering the fees they pay to accept high cost credit cards from the customers who choose to use them. c. The business-stealing effect is a negative consequence to merchants that surcharge credit card transactions: some customers will defect to nonsurcharging merchants. This will slow or deter the adoption of surcharging by some, but not all, merchants when merchant fees are at high levels, such as they are in the United States. 25 d. When cardholders switch to non-surcharged payments, credit card networks and Issuers lose transaction volume and fee revenue. Networks therefore have an 22 Separately, Professor Hausman argues that the state statutes which may restrict surcharging and American Express's policies with respect to surcharging reduce the potential effectiveness of surcharging. Hausman Report, 1154. I discuss those issues in Part 4.4. 23 Frankel Declaration, 1133. 24 25 Frankel Declaration, 1140 Frankel Declaration, 1136. 11

69511 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 15 of 85 PageID #: economic incentive to set merchant fees at lower levels if merchants can surcharge than if merchants cannot surcharge. 26 e. All merchants tend to benefit from the ability to surcharge, whether they surcharge or not, because of the constraining effect that merchants' collective ability to surcharge has on the level of interchange fees and network fees established by the card network. 27 Merchants also benefit to the extent that consumers' general preferences shift from using higher cost, more frequentlysurcharged credit cards, to lower cost payment cards and methods (e.g., debit cards). 28 f. To the extent that statutes in some states restrict merchants' ability to surcharge credit card transactions, that will reduce, but not eliminate, the benefits to merchants from gaining the ability to surcharge. 29 Similarly, to the extent that American Express's "non-discrimination" policy persists, that policy will reduce, but not eliminate, merchant benefits from the ability to surcharge credit card transactions. 3 20. In my Declaration, as with my earlier reports, I supported my analysis of the effects of relaxing the no-surcharge rules with available evidence from the record in this case and from jurisdictions where surcharging has been permitted. Professor Hausman agrees that surcharging can, in principle, benefit merchants and constrain interchange fee rates. 31 26 27 Frankel Declaration, Part 4.1.3. Frankel Declaration, 1144. 28 Frankel Declaration, 169. 29 Frankel Declaration, Part 4.6.1. 30 31 Frankel Declaration, Part 4.6.2. Hausman Report, 1076. Professor Hausman supported abolition of the no surcharge rule in New Zealand and reached conclusions similar to mine. See, e.g., Brief of Evidence of Professor Jerry Hausman, May 4, 2009, in Commerce Commission v. Cards NZ Limited et al., 115.2 ("Evidence from Australia demonstrates that some merchants will levy surcharges for credit card use to recover the MSFs when the 'no surcharge' rule is eliminated. If the no surcharge rules were relaxed, a significant number of consumers would switch to other lower cost payment vehicles, e.g. credit cards which did not have a surcharge. The result would be lower merchant costs since merchant acquirers would compete to keep their MSFs low so that their transactions would not be surcharged, and this business strategy would in turn place competitive pressure on issuers to keep any interchange fees low so that transactions using their cards would not be singled out for surcharge. Thus, an increase in competition would occur if the Visa and MasterCard no surcharge rules were eliminated..." "The least cost acceptance vehicle for many merchants is [debit cards]. The use of [debit cards] will increase if surcharges are levied on credit card transactions. Also, I would expect the usage of credit cards 12

69512 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 16 of 85 PageID #: However, in this case Professor Hausman disagrees with my interpretation of some of that evidence and concludes that surcharging in the United States is unlikely to occur or to be effective in eliminating Visa and MasterCard's market power. In the remainder of this Part, I review the main sources of evidence concerning the effectiveness of merchants' ability to surcharge and Professor Hausman's interpretation of that evidence. I explain that the evidence supports my conclusion that the ability to surcharge is beneficial to merchants. 4.1 Many Merchants Surcharge When Merchant Fees Are At Levels Prevailing For MasterCard And Visa Credit Card Acceptance In The United States 21. In my reports and Declaration, I concluded that economic incentives and common sense suggest that the likelihood that merchants will surcharge increases with the level of merchant fees that they pay to accept a particular brand of cards. 32 Evidence to support this conclusion included acknowledgements of this effect by MasterCard, 33 an Australian Bank, 34 Defendants' expert Professor Klein, 35 and Defendants' expert Professor Tope1. 36 In addition, I noted that surcharging and discounting were found in European merchant surveys to be more common in the Netherlands, where merchant fees were higher, with lower [merchant fees] to increase because surcharges on those cards would be lower that the surcharges on cards with higher [merchant fees]." 32 Frankel Declaration, 11]33; Frankel Report, 1111117, 184; Frankel Rebuttal Report, 11265. 33 34 35 36 "Payments System Regulation: Response by MasterCard Worldwide to the Issues for the 2007/08 Review," (Submission to the Reserve Bank of Australia), August 31, 2007, pp. 16-17 ("Merchants will have a higher incentive to surcharge the higher merchant fees are[...] An increase in merchant service fees will clearly raise the gains from surcharging relative to the costs, and hence make it more likely that surcharging will occur."). Letter of 31 August 2007 from Westpac Banking Corporation to Michelle Bullock, Head of Payments Policy Department, Reserve Bank of Australia, p. 2. Expert Report of Dr. Benjamin Klein, December 14, 2009, 111106 ("[A] higher interchange fee may increase merchants' incentives to steer cardholders away from using the payment card..."). Deposition of Robert H. Topel, Ph.D., April 20, 2010, p. 187. 13

69513 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 17 of 85 PageID #: than in Sweden, where merchant fees were lower. 37 In Australia, surcharging is more common for higher cost cards such as American Express than for lower cost cards such as MasterCard or Visa. 38 22. I also concluded that surcharging was likely to occur in the United States in the absence of rules prohibiting it if the Networks did not reduce the level of their interchange fees. The Networks' interchange fee rates in the United States are among the highest in the world, and substantially higher than any other country in which surcharging was permitted as of 2O08. also cited a study prepared by consultants for Visa that concluded that "[i]f the no- surcharge model was rescinded in the United States, it could lead to a significant reduction in interchange revenues... A large proportion of merchants would probably start surcharging." 4 4.1.1 In Australia, Merchants Commonly Surcharge Cards That Carry Merchant Fees Similar To The Networks' Cards In The United States 23. The experience with credit card reform in Australia, including the elimination of no-surcharge rules in that country, provides a particularly rich source of data and information. The Reserve Bank of Australia ("RBA") the country's central bank used its regulatory authority to permit merchants to surcharge credit card transactions beginning in 2003. The results support my conclusions that surcharging is more likely as the level of merchant fees 37 38 39 40 Frankel Declaration, 1133. East & Partners, Merchant Surcharging in Australia: Market Analysis Report Addendum for Friedman Law Group LLB [sic], March 2007, Table 2 ("a merchant who applies surcharges for a specific card type is three times more likely to surcharge AMEX than Diners cards and almost eight times more likely to surcharge AMEX compared to Visa cards.") Note that Diners is accepted by far fewer merchants than American Express. Expert Report of William Wecker, December 14, 2009, Figure 1. Frankel Rebuttal Report, 111266, citing OC&C Strategy Consultants, "Visa Litigation and the Changing Landscape A Time for Strategy," January 2006, VUSAMDL1-07908117. 14

69514 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 18 of 85 PageID #: increases and that surcharging is likely at the level of merchant fees prevailing in the United States. 24. At all times since 2003, the average cost to merchants in Australia to accept an American Express or Diners Club card has significantly exceeded the cost to merchants to accept a MasterCard or Visa credit card. The RBA also began to regulate MasterCard and Visa credit card interchange fee rates in 2003. Since the beginning of 2004, the cost to merchants to accept American Express cards has exceeded the cost of accepting MasterCard or Visa credit cards by an average of 1.2% of the purchase amount. 41 As a result, merchants in Australia have been far more likely to surcharge American Express card transactions than MasterCard and Visa credit card transactions. A survey conducted by East & Partners in Australia found that 39.1% of 2,243 merchants surveyed applied surcharges to at least some of their credit card transactions in June 2013. 42 Among the 33.3% of merchants that accepted American Express cards, 43 however, 78.1% applied surcharges to American Express cards. 44 A similar 79.2% of (far less numerous) merchants that accept Diners Club cards, another expensive brand, surcharge those cards. 45 I discuss Diners Club further in Part 4.3. 41 42 43 44 45 Reserve Bank of Australia Statistical Release C3, "Average Merchant Fees for Debit, Credit and Charge Cards," http://www.rba.gov.au/statistics/tables/xls/c03hist.xls (accessed June 22, 2013). Professor Hausman inaccurately describes the trend in this differential cost. Hausman Report, 1180. I discuss this trend further in Part 4.3. East & Partners, "Australian Merchant Payments: Market Analysis Report," June 2013 ("East & Partners, June 2013 Report"), Table 25. Id., Table 11. Id., Table 31. Id. 15

69515 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 19 of 85 PageID #: 25. American Express in Australia is a useful benchmark for the likelihood that merchants would want to surcharge MasterCard and Visa credit card transactions in the United States. That is because American Express has charged merchants an average of 2.11% of the transaction amount to accept American Express cards since December 2003. 46 Merchants pay even more than this roughly 2.34% in the United States to accept MasterCard and Visa credit cards (as of December 2009). 47 26. I noted in my Declaration that by December 2010, the RBA reported that 30% of Australian merchants overall were surcharging at least some credit card transactions. 48 By June 2012, the proportion of merchants surcharging credit card transactions had risen to 36% of surveyed merchants. 49 Figure 1 below updates Figure 1 in my Declaration. In addition, the 2013 update to the survey used by the RBA shows, as I noted above, that 39.1% of merchants now surcharge in Australia. Professor Hausman claims that my focus on the percentage of merchants that surcharge is "misguided" because it is inconsistent with a consumer panel study conducted by the RBA that found only 5% of participants' credit card transactions were 46 47 48 49 Reserve Bank of Australia Statistical Release C3, "Average Merchant Fees for Debit, Credit and Charge Cards," http://www.rba.gov.au/statistics/tables/xls/c03hist.xls (accessed June 22, 2013). By contrast, merchants have paid an average of only 0.91% in Australia to accept MasterCard and Visa credit card transactions, so it is unsurprising that they are less likely to surcharge those brands in Australia. Frankel Declaration, 1139. Frankel Declaration, $72, citing Reserve Bank of Australia, "Review of Card Surcharging: A Consultation Document," June 2011, p. 2. Even though MasterCard and Visa merchant service fees are substantially lower than American Express's, many merchants still surcharge them, although a smaller fraction of merchants accepting MasterCard and Visa surcharge those transactions than do merchants accepting American Express. This is consistent with my conclusion that higher rates to merchants are likely to result in higher levels of surcharging. Reserve Bank of Australia Payments System Board Annual Report, 2012, p. 25. 16

69516 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 20 of 85 PageID #: surcharged in 2007 and 2010, and, he claims, "it is the volume of transactions that are surcharged which provides the important economic factor." 5 Figure 1 Merchant Surcharging in Australia since No-Surcharge Rules Lifted in 2003 Very large merchants 45 30 Large merchants 45 30 15 Very small merchants 15 Small merchants 2006 2008 2010 2012 0 Source: Reserve Bank of Australia Payments System Board Annual Report, 2012, http://www.rba.gov.au/publications/annual-reports/psb/2012/pdf/2012-psb-ann-report.pdf, p. 25. 27. Professor Hausman is incorrect for two reasons. First, there is no inconsistency between finding that a large fraction of merchants surcharge and a much smaller percentage of credit card transactions are surcharged. As I already explained, merchants are far more likely to surcharge American Express transactions and Diners Club transactions (if they accept one or 50 Hausman Report, 1171. 17

69517 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 21 of 85 PageID #: both of those brands) than they are to surcharge the much less costly (in Australia) MasterCard and Visa credit cards. In December 2006, for example, a survey found that less than half of merchants that surcharged applied a surcharge to all credit cards; rather they disproportionately surcharged American Express cards. 51 American Express and Diners Club account for less than 20% of Australian credit and charge card transactions, with MasterCard and Visa accounting for the rest. 52 While all surveyed merchants in Australia accepted MasterCard and Visa credit cards in June 2013, only 33.3% of surveyed merchants accepted American Express cards (and only 6.9% accepted Diners Club cards). 53 28. Professor Hausman's second error is his related statement that it is the volume of transactions that are surcharged that is the "important economic factor." By surcharging only American Express cards (or American Express and Diners Club cards), however, merchants often induce their customers to switch to alternative, non-surcharged payment methods. This often will result in consumers using non-surcharged MasterCard and Visa credit cards. Finding that a much lower percentage of transactions are completed using surcharged credit cards is, in part, a reflection of the success of surcharging as a steering strategy. Inducing customers to use less costly payment methods that are not surcharged is one of the main purposes of surcharging. In addition, as I discuss further in Part 4.3, the mere threat to surcharge constrains the level of fees charged to merchants. Figure 1 shows that the likelihood of surcharging 51 East & Partners, Merchant Surcharging in Australia: Market Analysis Report Addendum for Friedman Law Group LLB [sic], March 2007, Table 2 ("a merchant who applies surcharges for a specific card type is three times more likely to surcharge AMEX than Diners cards and almost eight times more likely to surcharge AMEX compared to Visa cards.") Note that Diners is accepted by far fewer merchants than American Express. 52 http://www.rba.gov.au/statistics/tables/xls/c02hist.xls, accessed June 24, 2013. 53 East & Partners, June 2013 Report, Table 11. 18

69518 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 22 of 85 PageID #: increases with merchant size in Australia. Yet some large "strategic" merchants in Australia now receive extremely low interchange fee rates from MasterCard and Visa and do not surcharge. 54 To the extent they obtain these rates in part because they can credibly threaten to surcharge MasterCard and Visa transactions, this would also account for a relatively low percentage of transactions that are actually surcharged. 29. Professor Hausman suggests that the willingness of merchants to apply credit card surcharges in Australia does not shed light on the willingness of U.S. merchants to surcharge credit card transactions because, he asserts, retail competition is more intense in the United States than in Australia. He is apparently asserting that firms without market power are unlikely to surcharge. But 31.1% of "very small merchants" in Australia now surcharge. 55 In the United Kingdom, which Professor Hausman believes to be a useful benchmark for the United States, 56 the country's competition authority has concluded that the likelihood of surcharging is not related to a merchant's market power s' 30. While a merchant facing little competition may not suffer much from the business-stealing effect when surcharging, it is also true that in some highly competitive markets sellers will differentiate their prices to reflect differential costs. In the United Sates, retail gas stations have sometimes offered differential prices for credit card and cash 54 See, "Visa Interchange on Domestic Transactions in Australia," http://www.visa.com.au/ap/au/aboutvisa/interchange/interchange.shtml (visited August 9, 2013); "MasterCard domestic purchase transactions interchange fees for Australia," http://www.mastercard.com.au/merchant/getting started/interchange rates.html. 55 See Figure 1 and East & Partners, June 2013 Report, Table 25. 56 57 Hausman Report, 1l86. United Kingdom Office of Fair Trading, "Payment Surcharges: Response to the Which? Super-Complaint, July 2012 ("OFT Surcharging Report"), '115.19. 19

69519 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 23 of 85 PageID #: transactions due to the additional cost of accepting credit cards. The same has been true in Australia, leading to reduced interchange fees, as I discuss in Part 4.3. In addition, the intensity of competition varies among MasterCard and Visa "merchants" in the United States (which, for example, includes local governments, airlines, medical offices, commuter railroads, utilities, and many other types of sellers of goods and services in addition to retail stores like Wal-Mart). While business stealing concerns will deter some merchants from applying surcharges, others will likely surcharge, which lessens the business-stealing problem facing another merchant that elects to surcharge. 58 That is why the incidence of surcharging is likely to grow over time as it has in Australia. 59 4.1.2 Professor Hausman Misstates The Facts About The Level Of Interchange Fees and Merchant Fees In The United Kingdom 31. Professor Hausman argues that the experience with credit card surcharging in the United Kingdom, where merchants have been able to surcharge credit card transactions since 1992, "provides a better comparison" to what can be expected in the United States than does the experience of American Express in Australia. 60 He cites data which he interprets as 58 59 AmerExpMDL1720_0030683 708, at 695. AmerExpMDL1720_0030000, p. 2. 60 Hausman Report, 1186. 20

69520 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 24 of 85 PageID #: showing that surcharging is at "very low rates" in the United Kingdom despite "interchange rates... which remain quite high." 61 32. Professor Hausman,misstates the facts. First, the average Visa credit card interchange fee rate in the United Kingdom in 2007 a s in the United States. 62 Second, the number reported by Professor Hausman overstates the average merchant service fee in the U.K. Professor Hausman relies on a report from the United Kingdom's Office of Fair Trading ("OFT") the country's competition authority. The U.K. report estimates that the "illustrative costs of processing card transactions" for travel providers, i.e., the merchant service fee and other card transaction processing costs such as chargebacks and overhead, is 1.9% to 2.3%. 63 Professor Hausman finds that range similar to my estimate of the weighted average credit card merchant discount rate in the United States in 2009 of 2.34%. 64 The U.K. figures, however, include costs in addition to the merchant service fees, which the U.S. figures do not. The OFT Surcharging Report shows that the unweighted average merchant discount rate for retailers in the travel sector studied by the OFT i.e., excluding those additional costs not included in the U.S. estimate was 1.8%. 65 The report further notes that larger merchants tend to obtain lower card merchant discount rates than smaller merchants. 66 Thus, the weighted average merchant discount rate (typically referred to in the U.K. and 61 Id. 62 63 64 65 66 OFT Surcharging Report, pp. 69-70. Hausman Report, footnote 75. OFT Surcharging Report, p. 67. Id., p. 66. 21

69521 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 25 of 85 PageID #: Australia as the "merchant service charge") is lower than 1.8%. The OFT reports that the interchange fee in the U.K. typically accounts for about 70% of the merchant discount rate. With a weighted average interchange fee ofmas reported by Dr. Wecker, this suggests that the weighted average credit card merchant discount rate in the U.K. is roughly... all = 1.4%. This comports with a report prepared by U.K. consultants DotEcon for MasterCard in December 2003. 67 The DotEcon study includes a chart from a third-party source showing that, in 2002, the average merchant discount rate in the U.K. was about 1.5%. 68 This is well below the 2.34% that I estimated for the United States. 33. Furthermore, the U.K. figure cited by Professor Hausman contains other costs that do not appear in the U.S. figure, and so it is not an apples-to-apples comparison. Professor Hausman's figures, rather, include estimates made by the OFT of costs in addition to merchant discount rates as reported for the U.S., including average "charge back" costs, "overhead" costs, "intermediaries fees," and merchants' "own costs." 69 I did not include any such costs when estimating the weighted average U.S. credit card merchant discount rate. Thus, Professor Hausman is simply incorrect that merchant discount rates in the U.K. are similar to those in the 67 Economic Evidence in Support of MasterCard's Response to the Statement of Objections, December 2003 (MCI_MDL03_00015034). 68 Id., p. 14. DotEcon's own survey reports that the average merchant discount in the UK in 2003 was about 1.3% for the largest merchants and 2.0-2.5% for the smallest. Id., p. 63. 69 OFT Surcharging Report, 'ric.24 and Table C.3. The OFT was investigating how the level of surcharges compared to merchants' total costs of accepting credit card payments, and so added these additional cost elements. 22

69522 Case 1:05-md-01720-JG-JO Document 5939-5 Filed 08/16/13 Page 26 of 85 PageID #: United States, let alone that interchange fees in the U.K. are "quite high," particularly by U.S. standards. 7 34. The OFT Surcharging Report found that 14% of businesses in the U.K. applied surcharges to credit cards, and 9% applied surcharges to charge cards. 71 "Charge cards" consist of American Express and Diners Club, and East & Partners reports that only 10.1% of 505 surveyed merchants in the United Kingdom accepted American Express or Diners Club cards in 2009. 72 This suggests that a significant percentage of merchants that accept American Express or Diners Club cards in the U.K. apply surcharges to those cards. 73 35. The OFT concluded that "Netailers should still be able to impose transparent surcharges to consumers who choose to use payment mechanisms which cost more to process and offer discounts to consumers who choose to use payment mechanisms that cost less to process." 74 It explains: The OFT considers there are potentially benefits to both consumers and retailers if retailers are able to impose differential charges for different payment mechanisms. Surcharging or discounting different payment mechanisms can signal the costs of accepting each mechanism and therefore can help consumers make efficient choices between them. Consumers are then able to decide whether the benefits to them of paying by, for example, credit card... exceed the cost of paying by credit card. However, to achieve these signalling benefits, surcharges must be broadly in line with efficiently incurred costs. 70 71 72 Visa interchange rates for the United Kingdom in 2008 were low relative to other countries. See Wecker Report, Figure 1. OFT Surcharging Report, 115.5. East & Partners, "Australian and UK Credit Card Surcharging Perspectives: Custom Analysis for NERA Australia, November 2009, (HOUSTON001120), Table 16. 73 East & Partners' report on surcharging in the U.K. included only 51 merchants that accepted American Express or Diners Club in addition to MasterCard and Visa, 19 of which (37.3%) applied surcharges. Id., Tables 16, 19. That reported percentage is "for indicative purposes only" and not likely reliable as a specific percentage. 74 OFT Surcharging Report, 111.18. 23